Serious antisocial behavior (e.g., violence) peaks in adolescence and early adulthood and is most prevalent among persons whose prior history sows evidence of certain characteristics, conditions, and circumstances that constitute risk factors for such behavior. A promising approach to prevention is to intervene early in the developmental trajectory to alter those malleable risk factors presumed to mediate the development of antisocial behavior. To do this systematically, however, requires knowledge about (a) which risk factors at each age period are most strongly predictive of subsequent antisocial behavior and thus can be the basis for selecting "at risk" children for intervention; (b) the causal influence of the various risk factors on subsequent antisocial behavior so that intervention can be targeted on those factors most likely to actually have effects; and (c) which modes of intervention, for each age period, are most effective in favorably altering the targeted risk factors. A large volume of research bearing on each of these topics is already available but it has not yet been compiled and integrated in a way that fully reveals its implications for practice, policy, and theory. The project proposed here involves the use of systematic techniques of research synthesis (meta-analysis) to perform that task in a detailed and analytically rigorous fashion. It will (a) meta-analyze the research on the effects of intervention on risk factors for antisocial behavior and several bodies of related research, then (b) integrate the results with those of other relevant meta-analyses to construct a coherent account of current knowledge regarding risk-oriented intervention that is based on virtually all the pertinent empirical research available.